Thursday, August 5, 2010 - 8:20 AM

COS 87-2: Exploitation of host plant morphospace by generalist and specialist phytophagous insects

Stephen B. Heard, University of New Brunswick and William Godsoe, University of Tennessee.

Background/Question/Methods

Dozens of phytophagous insect species feed on the related goldenrods Solidago altissima and S. gigantea, with some feeding as relative generalists and others having evolved complexes of host-specialist races or cryptic species.  Insects exploiting such a host-species pair must make decisions about which individual plants to attack, and will also experience differential performance across individual plants.  Patterns in either preference or performance can lead to differential herbivore attack between host species but also among plant morphotypes within host species. 

We examined patterns of attack by 10 insect herbivores within and between goldenrod species, with respect to a plant morphospace defined by seven traits plausibly related to host choice or performance.  For each herbivore on each goldenrod species, we asked whether individual plants experiencing attack lay in a nonrandom subset of plant morphospace, and if so, if that subset was central or marginal to available host morphospace.  Where herbivores on at least one goldenrod species attacked a subset of available host morphospace, we also asked whether attacked morphotypes of S. altissima and S. gigantea were more similar or more different than expected at random.  Models of host-associated diversification suggest that the results of analyses like ours might differ between generalists and pairs of host specialists, and between younger and older pairs of specialists (with young pairs of specialists exploiting distant subsets of overall host morphospace).

Results/Conclusions

For 8 of 10 insect species or cryptic species pairs, we were unable to reject the hypothesis of exploitation of the complete host-plant morphospace.  However, for a leafminer and an aphid attacked plants appeared to define a subset of available morphospace on each host.  The relationship between the two morphospace subsets differed between insect species: for the leafminer the two morphospace subsets were close together (miners attacked altissima-like S. gigantea and gigantea-like S. altissima), but for an aphid the two morphospace subsets were defined by an axis of variation orthogonal to that expressing variation between the hosts.   Similar data for known pairs of young host-specialists are not yet available, but are needed to test our understanding of ecological controls on host-associated diversification.